No abstract

This paper describes a computationally-enhanced M100 UAV platform with an onboard deep learning inference system for integrated computer vision and navigation. The system is able to autonomously find and visually identify by coat pattern individual Holstein Friesian cattle in freely moving herds. We propose an approach that utilises three deep convolutional neural network architectures running live onboard the aircraft: (1) a YOLOv2-based species detector, (2) a dual-stream deep network delivering exploratory agency, and (3) an InceptionV3-based biometric long-term recurrent convolutional network for individual animal identification. We evaluate the performance of each of the components offline, and also online via real-world field tests comprising 147 minutes of autonomous low altitude flight in a farm environment over a dispersed herd of 17 heifer dairy cows. We report errorfree identification performance on this online experiment. The presented proof-of-concept system is the first of its kind. It represents a practical step towards autonomous biometric identification of individual animals from the air in open pasture environments for tag-less AI support in farming and ecology.

A UAV with a variable sweep wing has the potential to perform a perched landing on the ground by achieving high pitch rates to take advantage of dynamic stall. This study focuses on the generation and evaluation of a trajectory to perform a perched landing on the ground using a non-linear constraint optimiser (Interior Point OPTimizer) and a Deep Q-Network (DQN). The trajectory is generated using a numerical model that characterises the dynamics of a UAV with a variable sweep wing which was developed through wind tunnel testing. The trajectories generated by a DQN have been compared with those produced by non-linear constraint optimisation in simulation and flown on the UAV to evaluate performance. The results show that a DQN generates trajectories with a lower cost function and have the potential to generate trajectories from a range of starting conditions (on average generating a trajectory takes 174 milliseconds). The trajectories generated performed a rapid pitch up before the landing site is reached, to reduce the airspeed (on average less than 0.5m/s just above the landing site) without generating an increase in altitude, and then the nose dropped just before hitting the ground to allow the aircraft to be recovered without damaging the tail. The trajectories generated by a DQN produced a final airspeed (when it hit the ground)

This paper describes a series of proof-of-concept Beyond Visual Line Of Sight unmanned aerial vehicle flights which reached a range of up to 9 km and an altitude of 4,410 m Above Mean Sea Level over Volcán de Fuego in Guatemala, interacting with the volcanic plume on multiple occasions across a range of different conditions.Volcán de Fuego is an active volcano which emits gas and ash regularly, causing disruption to airlines operating from the international airport 50 km away and impacting the lives of the local population. Collection of data from within the plume develops scientists' understanding of the composition of the volcano's output and is of use to scientists, aviation, and hazard management groups alike. This paper presents preliminary results of multiple plume interceptions with multiple aircraft, carrying a variety of sensors. A plume-detection metric is introduced, which uses a combination of flight data and atmospheric sensor data to identify flight through a volcanic plume. Future work will develop the automation of plume tracking such that reliable scientific data sets can be gathered in a robust manner.

As part of an NERC-funded project investigating the southern methane anomaly, a team drawn from the Universities of Bristol, Birmingham and Royal Holloway flew small unmanned multirotors from Ascension Island for the purposes of atmospheric sampling. The objective of these flights was to collect air samples from below, within and above a persistent atmospheric feature, the Trade Wind Inversion, in order to characterise methane concentrations and their isotopic composition. These parameters allow the methane in the different air masses to be tied to different source locations, which can be further analysed using back trajectory atmospheric computer modelling. This paper describes the campaigns as a whole including the design of the bespoke eight rotor aircraft and the operational requirements that were needed in order to collect targeted multiple air samples up to 2.5 km above the ground level in under 20 min of flight time. Key features of the system described include real-time feedback of temperature and humidity, as well as system health data. This enabled detailed targeting of the air sampling design to be realised and planned during the flight mission on the downward leg, a capability that is invaluable in the presence of uncertainty in the pre-flight meteorological data. Environmental considerations are also outlined together with the flight plans that were created in order to rapidly fly vertical transects of the atmosphere whilst encountering changing wind conditions. Two sampling campaigns were carried out in September 2014 and July 2015 with over one hundred high altitude sampling missions. Lessons learned are given throughout, including those associated with operating in the testing environment encountered on Ascension Island.

A new method for enabling a quadrotor micro air vehicle (MAV) to navigate unknown environments using reinforcement learning (RL) and model predictive control (MPC) is developed. An efficient implementation of MPC provides vehicle control and obstacle avoidance. RL is used to guide the MAV through complex environments where dead-end corridors may be encountered and backtracking is necessary. All of the presented algorithms were deployed on embedded hardware using automatic code generation from Simulink. Results are given for flight tests, demonstrating that the algorithms perform well with modest computing requirements and robust navigation. Keywords Model predictive control • Reinforcement learning • Exploration • Micro air vehicle 1 Introduction This paper introduces a method for navigation and control of quadrotors within a non-convex obstacle field. The method uses online optimization within a model predictive control (MPC) framework, taking advantage of Fast MPC (Wang and Boyd 2010) with soft constraint modifications (Richards 2015) to provide a real-time controller on embedded hardware. Furthermore, the use of reinforcement learning (RL) enables autonomous navigation by providing high level path planning decisions for navigation of previously unexplored spaces. Flight test experiments demonstrate the methods within a two dimensional control scenario. The experiments use off-board localization by motion capture and synthesized sensing of obstacles: although these also include important challenges, the focus here is on the decision-making. Trajectory generation in the presence of obstacles is NP-hard (Reif 1979) and has been the subject of considerable algorithm development, including randomized B Arthur G. Richards

No abstract

A variable sweep wing UAV is developed utilising off the shelf components with a custom mechanism for the wing box. The movement of the wing sweep in flight enables large pitching moments suitable for performing perching manoeuvres. Wind tunnel data is presented that confirms the favourable characteristics expected from sweeping the wing and achieving high pitch rates. Whilst only small sweep changes are required during flight, the design allows up to 30 • forward sweep for significant pitching moments during the flare. A new collection of controllers is developed based on observations from similar landing techniques performed by birds and hang-gliders onto flat ground. The three-stage landing process takes the aircraft along an approach path, through a roundout procedure during which airspeed decays and concludes with rapid pitch up. Flight test results are presented during which it is found that the airspeed can be reduced to, on average, under 3m/s in the final moments before landing-well below the stall speed of 9m/s.